Skip to main content
NCBI home page
Clinical Orthopaedics and Related Research logoLink to Clinical Orthopaedics and Related Research
. 2020 Aug 10;478(12):2921–2933. doi: 10.1097/CORR.0000000000001422

Is Use of BMP-2 Associated with Tumor Growth and Osteoblastic Differentiation in Murine Models of Osteosarcoma?

Joseph K Kendal 1,2,3,4, Arvind Singla 1,2,3,4, Asmaa Affan 1,2,3,4, Kurt Hildebrand 1,2,3,4, Abdullah Al-Ani 1,2,3,4, Mark Ungrin 1,2,3,4, Douglas J Mahoney 1,2,3,4, Doha Itani 1,2,3,4, Frank R Jirik 1,2,3,4, Michael J Monument 1,2,3,4,✉
PMCID: PMC7899403  PMID: 33275394

Abstract

Background

The putative benefit of rhBMP-2 is in the setting of limb reconstruction using structural allografts, whether it be allograft-prosthetic composites, osteoarticular allografts, or intercalary segmental grafts. There are also potential advantages in augmenting osseointegration of uncemented endoprosthetics and in reducing infection. Recombinant human BMP-2 might mitigate nonunion in structural allograft augmented osteosarcoma limb salvage surgery; however, its use is limited because of concerns about the prooncogenic effects of the agent.

Questions/purposes

(1) To assess if BMP-2 signaling influences osteosarcoma cell line growth. (2) To characterize degree of osteosarcoma cell line osteoblastic differentiation in response to BMP-2. (3) To assess if BMP-2 signaling has a consistent effect on local or systemic tumor burden in various orthotopic murine models of osteosarcoma.

Methods

In this study, 143b, SaOS-2 and DLM8-M1 osteosarcoma cell lines were transfected with BMP-2 cDNA controlled by a constitutive promoter (experimental) or an empty vector (control) using a PiggyBac transposon system. Cellular proliferation was assessed using a quantitative MTT colorimetric assay. Osteoblastic differentiation was compared between control and experimental cell lines using quantitative real-time polymerase chain reaction of the osteoblastic markers connective tissue growth factor, Runx-2, Osterix, alkaline phosphatase and osteocalcin. Experimental and control cell lines were injected into the proximal tibia of either NOD-SCID (143b and SaOS-2 xenograft model), or C3H (DLM8-M1 syngeneic model) mice. Local tumor burden was quantitatively assessed using tumor volume caliper measurements and bioluminescence, and qualitatively assessed using post-mortem ex vivo microCT. Lung metastasis was qualitatively assessed by the presence of bioluminescence, and incidence was confirmed using histology. rhBMP-2 soaked absorbable collagen sponges (experimental) and sterile-H2O soaked absorbable collagen sponges (control) were implanted adjacent to 143b proximal tibial cell line injections to compare the effects of exogenous BMP-2 application with endogenous upregulation.

Results

Constitutive expression of BMP-2 increased the in vitro proliferation of 143b cells (absorbance values 1.2 ± 0.1 versus 0.89 ± 0.1, mean difference 0.36 [95% CI 0.12 to 0.6]; p = 0.01), but had no effect on SaOS-2 and DLM8-M1 cell proliferation. In response to constitutive BMP-2 expression, 143b cells had no differences in osteoblastic differentiation, while DLM8-M1 cells downregulated the early marker connective tissue growth factor (mean ΔCt 0.2 ± 0.1 versus 0.6 ± 0.1; p = 0.002) and upregulated the early-mid range marker Runx-2 (mean ΔCt -0.8 ± 0.1 versus -1.1 ± 0.1; p = 0.002), and SaOS-2 cells upregulated the mid-range marker Osterix (mean ΔCt -2.1 ± 0.6 versus -3.9 ± 0.6; p = 0.002). Constitutive expression of BMP-2 resulted in greater 143b and DLM8-M1 local tumor volume (143b: 307.2 ± 106.8 mm3 versus 1316 ± 387.4 mm3, mean difference 1009 mm3 [95% CI 674.5 to 1343]; p < 0.001, DLM8-M1 week four: 0 mm3 versus 326.1 ± 72.8 mm3, mean difference 326.1 mm3 [95% CI 121.2 to 531]; p = 0.009), but modestly reduced local tumor growth in SaOS-2 (9.5 x 108 ± 8.3x108 photons/s versus 9.3 x 107 ± 1.5 x 108 photons/s, mean difference 8.6 x 108 photons/s [95% CI 5.1 x 108 to 1.2 x 109]; p < 0.001). Application of exogenous rhBMP-2 also increased 143b local tumor volume (495 ± 91.9 mm3 versus 1335 ± 102.7 mm3, mean difference 840.3 mm3 [95% CI 671.7 to 1009]; p < 0.001). Incidence of lung metastases was not different between experimental or control groups for all experimental conditions.

Conclusions

As demonstrated by others, ectopic BMP-2 signaling has unpredictable effects on local tumor proliferation in murine models of osteosarcoma and does not consistently result in osteosarcoma cell line differentiation. Further investigations into other methods of safe bone and soft tissue healing augmentation and the use of differentiation therapies is warranted.

Clinical Relevance

Our results indicate that BMP-2 has the potential to stimulate the growth of osteosarcoma cells that are poorly responsive to BMP-2 mediated osteoblastic differentiation. As this differentiation potential is unpredictable in the clinical setting, BMP-2 may promote the growth of microscopic residual tumor burden after resection. Our study provides further support for the recommendation to avoid the use of BMP-2 after limb-salvage surgery in patients with osteosarcoma.

Introduction

Osteosarcoma is the most common primary malignant bone tumor and predominantly affects children and young adults. Curative surgery involves aggressive resection of bone, muscle, and articular structures. Limb salvage is possible in more than 85% of patients; however, limb reconstruction necessitates the use of endoprosthetic implants, structural bone grafts, or both [13]. Bone healing is essential to the success of biologic limb-salvage reconstruction, but postoperative soft-tissue deficits and cytotoxic chemotherapy substantially impair healing potential. Structural bone allograft reconstructions are preferred when the tumor resection does not involve the joint; however, they are associated with a staggering 20% to 40% complication proportion and a 60% proportion of surgical revision because of delayed or absent bone healing, implant loosening and fracture, allograft fracture, and infection [19]. In endoprosthetic reconstructions, impaired bone healing compromises osseointegration into the metallic implant and diminishes hypertrophy around the bone-implant interface, contributing to aseptic loosening [16]. Surgical adjuncts to improve early bone healing in patients with osteosarcoma are needed to improve surgical and patient outcomes, reduce revision surgery, and reduce the cost of limb-salvage treatments.

BMP-2 is a potent proinflammatory stimulator of bone healing and connective tissue repair [15]. BMP-2 induces bone formation via recruitment and proliferation of mesenchymal progenitor cells and induction of osteoblast lineage differentiation and terminal osteoblastic differentiation of committed cells [7, 23]. Recombinant human BMP-2 (rhBMP-2) delivered on an absorbable collagen sponge is currently approved by the FDA and Health Canada as a surgical adjunct to promote bone healing after orthopaedic surgery [3].

Adding rhBMP-2 on an absorbable collagen sponge to structural allograft based limb-salvage surgery for osteosarcoma could be a simple and effective strategy to improve bone healing and protect against infection [12]. However, concerns about unwanted upregulation of prooncogenic cell signaling pathways in the resection bed of a mesenchymal cell-derived malignancy has limited the application of rhBMP-2 on an absorbable collagen sponge during surgery to treat bone sarcomas. Evidence supporting this concern is limited and controversial. Recent studies have demonstrated that BMP-2 does not alter the proliferation, migration, or invasion of osteosarcoma cells [11], but may promote differentiation of primitive osteosarcoma cells [27]. In two recent xenograft studies, application of rhBMP-2 to amputation sites in patients with osteosarcoma did not increase the local recurrence of osteosarcoma [8], and simultaneous tail-vein injections of osteosarcoma cells and rhBMP-2 did not increase lung metastases [11].

We therefore aimed to (1) assess if BMP-2 signaling influences osteosarcoma cell line growth; (2) characterize the degree of osteosarcoma cell line osteoblastic differentiation in response to BMP-2; and (3) assess if BMP-2 signaling has a consistent effect on local or systemic tumor burden in various orthotopic murine models of osteosarcoma.

Materials and Methods

Experimental Overview

To assess the effects of BMP-2 on osteosarcoma cell proliferation, osteoblastic phenotype and tumor burden, we conducted a series of in vitro and in vivo experiments (Fig. 1). Three osteosarcoma cell lines with varying degrees of osteoblastic differentiation (143b, SaOS-2 and DLM8-M1) were genetically engineered to constitutively express BMP-2 (experimental condition). Controls included cell lines with native BMP-2 expression (empty vector transfection). We assessed cell proliferation and two-dimensional (2-D) invasion using an MTT assay and scratch assay, respectively. We used quantitative reverse transcription polymerase chain reaction (qRT-PCR) with a panel of markers along the spectrum of osteoblastic differentiation to assess if constitutive BMP-2 expression altered the differentiation profile of osteosarcoma cell lines. In vivo modelling involved both xenograft and syngeneic orthotopic osteosarcoma murine models. The syngeneic model allowed assessment of the effects of BMP-2 in the setting of an intact murine immune system, which is important given the proinflammatory effects of BMP-2. Control and experimental cell lines were injected orthotopically into the proximal tibiae of either NOD-SCID (xenograft model) or C3H (syngeneic model) mice. Tumor burden was assessed in vivo using bioluminescence and tumor volume measurements, and ex vivo using microCT and histopathology (Fig. 1). Lung metastases were identified as present or absent using bioluminescence in vivo and histopathology postmortem. An in vivo experiment with the 143b cell line (without BMP-2 constitutive expression) was also performed assessing the effects of local application of either a rhBMP-2 soaked absorbable collagen sponge (experimental condition) or a sterile H2O soaked absorbable collagen sponge (control condition).

Fig. 1.

Fig. 1

Open in a new tab

This figure provides an overview of the experiment. Osteosarcoma cell lines (SaOS-2, 143b and DLM8-M1) were engineered to constitutively express BMP-2 (experimental cell lines). Control cell lines were transfected with an empty vector. In vitro studies included an MTT assay (proliferation assay), scratch assay (wound healing assay) and quantitative reverse transcription polymerase chain reaction (qRT-PCR) using various markers of osteoblastic differentiation. Intratibial injections of osteosarcoma cell lines were performed, followed by an in vivo assessment of tumor burden using tumor volume measurements and bioluminescence. Ex vivo assessments performed included microCT of primary tumors and histologic assessment of primary tumors and lungs; CTGF = connective tissue growth factor, Osx = Osterix, ALP = alkaline phosphatase, OC = osteocalcin.

Cell Line Characteristics and Maintenance

Three different osteosarcoma cell lines were used to represent phenotypic diversity and assess the effects of BMP-2 across multiple osteosarcoma models. We obtained 143b and SaOS-2 cells from the American Type Culture Collection. The 143b cell line derives from the TE-85 HOS cell line obtained from a 13-year-old white female. The 143b cell line was created in 1975, when the K-ras oncogene was introduced into TE-85 cells using the Kirsten mouse sarcoma virus [22]. The 143b cells have high proliferative capacity and a greater potential for pulmonary metastasis, and produce extensive osteolytic lesions [30]. SaOS-2 cells have a lower proliferative and metastatic profile and produce a dense osteoid matrix. SaOS-2 cells were originally derived from a primary osteosarcoma from an 11-year-old white female. The mouse osteosarcoma cell line DLM8-M1 was obtained via a material transfer agreement from Dr. Nicole Ehrhart’s laboratory at Colorado State University (Fort Collins, CO, USA) [1]. DLM8-M1 is derived from a C3H mouse background. LM8 cells were originally derived from a Dunn OS cell line that was cultured from a spontaneous osteosarcoma in a C3H mouse [2].

We maintained 143b cells in RPMI-1640 with 10% fetal bovine serum. SaOS-2 cells were maintained in McCoy’s 5A (modified) media with 15% fetal bovine serum. DLM8-M1 cells were maintained in modified Dulbecco’s modified eagle medium with 4.5 g/L of D-glucose and 10% fetal bovine serum. We added 100 U/mL (1%) of penicillin and 100 μg/mL (1%) of streptomycin to all media. Cells were maintained in a sterile incubator at 5% CO2 and 37o C.

Luciferase Marker Transfection

All cell lines were engineered to express the bioluminescent protein, Luciferase, to allow for tumor burden quantification by capturing weekly photon counts in vivo. We transfected SaOS-2 cells with a dual-reporter plasmid vector for enhanced GFP and firefly luciferase fusion protein [4, 25]. We infected 143b cells with a dual-reporter lentivirus encoding mCherry and luciferase (pLV430G-oFL-T2A-mCherry lentivirus). Incorporation of pLV430G-oFL-T2A-mCherry lentiviral was initially confirmed using fluorescence microscopy. DLM8-M1 cells were infected with the pLV430G-oFL-T2A-mCherry retrovirus. These cells were then expanded in culture and FACS-sorted using mCherry as the marker. All cell lines were negative for mycoplasma contamination (Centre for Genome Engineering, University of Calgary (Calgary, AB, Canada), last tested March 12, 2018).

BMP-2 cDNA Transfection

To assess for the phenotypic consequences of continuous BMP-2 signaling, we engineered all three cell lines to constitutively express BMP-2 homodimer. Human BMP-2 cDNA was obtained in a plasmid (pGEM-BMP2) from Sino Biological Inc (Shanghai, China). A PiggyBac transposon system was used (PB-CMV-MCS-EF1α-Puro plasmid, System Biosciences, Palo Alto, CA, USA). Human BMP-2 cDNA was subcloned from the pGEM-BMP2 vector to the PB-CMV-MCS-EF1α-Puro plasmid generating the PBac B.M.P (BMP2-MCS-Puro) plasmid (see Fig. 1; Supplemental Digital Content 1, http://links.lww.com/CORR/A396).

Stable transfection proceeded with the use of the Super PiggyBac Transposase vector, combined with 8 μL of lipofectamine 2000 reagent and 50 μL of Opti-MEM media. We performed puromycin titrations. Positive control cell lines were generated using the stock PB-CMV-MCS-EF1α-Puro plasmid. Transfected cells were grown to 90% to 100% confluency. We used 50 μL of culture supernatant to assess the expression of BMP-2 via an enzyme-linked immunosorbent assay (BMP-2 Quantikine Enzyme-linked Immunosorbent Assay, R&D Systems, Minneapolis, MN, USA) [21].

MTT Assay

To assess for the effects of constitutive BMP-2 expression on osteosarcoma cell line proliferation, we performed an MTT assay. Control and experimental cell lines were plated in triplicate. After the cells were incubated for 24 hours, 48 hours, 72 hours, and 96 hours, we added 10 μL of MTT (50 mg/mL) reagent. Plates were incubated for 4 hours. We added 100 μL of isopropanol with 0.04N HCl to each well. Absorbance was tested after 1 hour at 550 nm. Assessors were aware of the plate conditions.

Scratch Assay (Wound Healing Assay)

To assess the effects of constitutive BMP-2 expression on the 2-D motility of osteosarcoma cell lines, we conducted a scratch assay. Cell lines were plated in triplicate in six-well plates and grown to 100% confluency. Opti-MEM media was added the night before for serum starvation. A wound was created using a 200-μL pipette tip. Images were captured at 0 hours, 4 hours, 8 hours, 12 hours, and 24 hours after scratching. Photographs were taken under inverted microscopy using QCapture software (Teledyne QImaging, Surrey, BC, Canada) at 50x magnification. The scratch width was subsequently quantified in pixels (QCapture, Teledyne QImaging). Time 0 was used as a reference point, indicating 100% scratch width, and subsequent timepoints were measured as a percentage of the baseline scratch width. Assessors were aware of the plate conditions.

Primer Design, RNA Extraction, cDNA Synthesis, and Quantitative Real-time polymerase Chain Reaction

To assess how constitutive BMP-2 signaling affected degree of osteoblastic differentiation in osteosarcoma cell lines, we used qRT-PCR using osteoblastic markers along the spectrum of differentiation. Primers were designed for connective tissue growth factor, alkaline phosphatase, Osterix, Runx-2, BMP-2, and osteocalcin (see Table 1; Supplemental Digital Content 2, http://links.lww.com/CORR/A397). Primer sequences were either derived from previous studies or designed de novo. Primer sets were generated for Homo sapiens and Mus musculus samples.

RNA was extracted from frozen cell pellets (Norgen Biotek, Thorold, ON, Canada). iScript Reverse Transcription Supermix (Bio-Rad Laboratories, Mississauga, ON, Canada) was used to generate cDNA. Quantitative PCR (qPCR) was performed using a fluorescent dye detection method (“BrightGreen,” Applied Biological Material Inc, Richmond, BC, Canada). We completed qPCR reactions in technical duplicates by an automated pipettor (epMotion 5070, Eppendorf, ON, Canada). The QuantStudio 6 Flex Real-time PCR platform (ThermoFisher Scientific, Waltham, MA, USA) was used for data analysis. Obtained Ct values were compared with those of an optimized housekeeping gene (TATA-binding protein for DLM8-M1; G6PD for SaOS-2 and 143b). A Mann-Whitney U-test was performed to statistically analyze resultant ΔCt values.

Animal Tumor Models

Animal studies were performed in compliance with the Canadian Council on Animal Care guidelines, with ethics approval from the University of Calgary Animal Care Committee (AC17-0165). For the xenograft models, 6- to 8-week-old male and female NOD.CB17-Prkdcscid/J (NOD-SCID) mice (Charles River Laboratories, Senneville, Quebec, Canada) were used. For the syngeneic model, 6-week-old female C3HeB/FeJ (“C3H”) (Charles River Laboratories) were used. A breeding protocol was established for the NOD SCID mice (obtained from Charles River originally), therefore, we used both male and female mice. C3H mice were ordered, so we chose to control the sex. Mice were anesthetized using a combination of xylazine (10 mg/kg) and ketamine (100 mg/kg) via an intraperitoneal injection. A cortical defect was created using a 26-gauge needle directed through the proximal tibial metaphysis. We injected 10 μL of viable cell suspension (1 x 105 for 143b cells, 5 x 105 for SaOS-2, or 5 x 105 for DLM8-M1 cells in serum-free media) over 60 seconds into the proximal tibia’s intramedullary space using a 30-gauge Hamilton syringe (Hamilton Company, Reno, NV, USA). Meloxicam was administered subcutaneously at a dose of 2 mg/kg for postinjection analgesia. We randomly assigned the animals to the control or experimental groups and when applicable, males and females were equally divided between groups to minimize the variable of sex. Cages were frequently monitored for adverse events by the animal housing staff and the attending veterinarian. Mice were euthanized at the first sign of distress, including evidence of piloerection, respiratory distress, weight loss, reduced or altered mobility and stance, hunched posture or poor grooming, infection unresponsive to therapy or tumor size > 1.5 cm in diameter. Tumor size humane endpoint was reached at 4 weeks for BMP-2 stimulated 143b derived tumors, and at 5 weeks for BMP-2-overexpressing DLM8-M1 derived tumors. One mouse in each control and experimental SaOS-2 cohorts had off-target injections (soft-tissue) and were excluded. One mouse in the SaOS-2 experimental cohort was found deceased at week 6 and one mouse in the DLM8-M1 experimental cohort was found deceased at week 5. One mouse in the SaOS-2 experimental cohort developed lymphoma and was excluded.

To assess the effects of exogenous rhBMP-2 on an absorbable collagen sponge, we soaked 10 μg of rhBMP-2 (PeproTech Canada, Cranbury, NJ, USA) on a sterile absorbable collagen sponge and then surgically implanted it into the submuscular tibial space immediately before 143b cell line injection (see Fig. 2A-C; Supplemental Digital Content 3, http://links.lww.com/CORR/A398) [21, 29]. Negative controls were administered 143b cell line injections with implantation of a sterile absorbable collagen sponge soaked in autoclaved H2O. To assess effects of the endogenous overexpression of BMP-2 on local and systemic tumor burden, we prepared and injected cells (143b, SaOS-2 and DLM8-M1 cell lines) overexpressing BMP-2 and positive control cells as described.

Bioluminescence was performed weekly until the tumor size reached at least 1.5 cm in the greatest dimension, or until 6 weeks after injection. Bioluminescence values were normalized to background emission. Mice were administered 150 mg/kg of D-luciferin via an intraperitoneal injection and anesthetized using isoflurane (1.5% v/v% in 1 L/min O2). Imaging was performed with the Xenogen IVIS Lumina system (Caliper Life Sciences, Waltham, MA, USA). Images were captured 10 minutes after the injection. The tumor’s volume was measured manually with calipers and calculated using the formula (L + W) (L) (W) (0.2618) [18]. Assessors were aware that the cages were labeled “experimental” or “control.”

Mice were euthanized using continuous carbon dioxide inhalation followed by cervical dislocation. Necropsy was performed. Hindlimb and lung samples were fixed in 10% neutral buffered formalin overnight (VWR International, Radnor, PA, USA). After microCT scans were performed, the limbs were prepared for a histopathologic assessment. Formalin-fixed samples were decalcified using Immunocal and embedded in paraffin blocks. Samples slices were stained with hematoxylin and eosin for microscopic analysis. All histologic specimens were reviewed and validated by a clinical sarcoma pathologist (DI).

MicroCT

An ex vivo microCT analysis was performed using Scanco vivaCT 40 (Scanco Medical AG, Basserdorf, Switzerland). Scans were performed with an isotropic resolution of 15 μm at a tube voltage of 45 kVp, current of 133 μA, and integration time of 200 ms. Data from each scan were extracted via the Image Processing Language software (V5.08b, Scanco Medical, Brüttisellen, Switzerland). Three-dimensional (3-D) reconstructions and axial slices were analyzed for each hindlimb. For SaOS-2-derived tumors, a region of interest for analysis was designated at the first 100 slices distal to the anterior physis. Contours were generated for analysis to provide separate assessments of the cortical and trabecular bone. Extracted datapoints included the trabecular thickness, bone mineral density (assessed via gray-scale values), trabecular connectivity density, and trabecular number. Contralateral hindlimbs were scanned for an internal control of baseline values, and the percentage change in values from baseline were then compared between control and experimental cohorts.

Statistical Analysis

Continuous data was assessed using an unpaired student’s t-test. For categorical variables and multiple group comparisons where each data point was dependent on the previous data point, a two-way ANOVA was used. For bioluminescence data, a post-hoc analysis was performed using Sidak’s multiple comparison test. A p value of < 0.05 was considered to be of statistical significance. GraphPad Prism 8 (GraphPad Software, San Diego, CA) was used for both statistical analysis and generation of graphs. In all figures, corresponding letters represent statistical significance as follows: ap < 0.05; bp < 0.01; cp < 0.001; and dp < 0.0001.

Results

BMP-2 Overexpression Shows a Variable Effect on Cell Proliferation and Migration in Human and Murine Osteosarcoma Cell Lines

We found that 143b, SaOS-2, and DLM8-M1 experimental cell lines overexpressed the BMP-2 homodimer as assessed by the enzyme-linked immunosorbent assay (see Fig. 3; Supplemental Digital Content 4, http://links.lww.com/CORR/A399). The mean concentration of 143b control cell culture supernatant was 50.1 ± 3.6 pg/mL versus 272.7 ± 21.7 pg/mL for 143b experimental cells (mean difference 222.2 [95% CI 187.0 to 257.4]; p < 0.01). SaOS-2 control cells produced a mean concentration of 51 ± 4.9 pg/mL versus 349.4 ± 14.2 pg/mL of BMP-2 produced by SaOS-2 experimental cells (mean difference 298.4 [95% CI 274.3 to 322.5]; p < 0.01). In DLM8-M1 control cells, the mean BMP-2 concentration was 38 ± 30 pg/mL versus 287.6 ± 32.3 pg/mL in DLM8-M1 experimental cells (mean difference 249.8 [95% CI 179.2 to 320.5]; p < 0.01).

In the MTT assay, BMP-2 overexpression stimulated 143b cell line proliferation but had no effect on SaOS-2 or DLM8-M1 cell lines. 143b experimental cells had greater absorbance than control cells at day 4 (1.2 ± 0.1 versus 0.89 ± 0.1, mean difference 0.36 [95% CI 0.12 to 0.6]; p = 0.01) (see Supplemental Digital Content 5, http://links.lww.com/CORR/A400). No differences in absorbance were identified at all points between experimental and control cell lines for SaOS-2 (see Fig. 4B; Supplemental Digital Content 6, http://links.lww.com/CORR/A401) and DLM8-M1 cells lines (see Fig. 4C; Supplemental Digital Content 7, http://links.lww.com/CORR/A402).

143b experimental cells demonstrated an increase in scratch wound closure compared with control cells at 8 hours (mean difference 23% [95% CI 5.9 to 40.1]; p < 0.01) and 24 hours (mean difference 23% [95% CI 6.0 to 40.1]; p < 0.01) (see Fig. 5A; Supplemental Digital Content 8, http://links.lww.com/CORR/A403). In SaOS-2 cells and DLM8-M1, there was no difference in scratch width over time (see Fig. 5B-C; Supplemental Digital Content 9, http://links.lww.com/CORR/A404 and Supplemental Digital Content 10, http://links.lww.com/CORR/A405). These results indicate that BMP-2 overexpression had no effect on 2-D motility in SaOS-2 or DLM8-M1 cells but increased motility in the 143b cell line.

BMP-2 Induction of Osteoblastic Differentiation Markers Differs Between the Osteosarcoma Cell Lines

A comparison of gene expression was performed to assess SaOS-2 and 143b using the same housekeeping gene as a reference (G6PD). SaOS-2 had an increased expression of Runx-2 (mean ΔCt -1.5 ± 1.0 versus 0.2 ± 0.5; p = 0.009), Osterix (mean ΔCt -2.1 ± 0.6 versus 2.2 ± 0.5; p = 0.004), and alkaline phosphatase (mean ΔCt -4.2 ± 0.4 versus 4.1 ± 0.7; p = 0.04), while SaOS-2 had decreased expression of connective tissue growth factor (mean ΔCt -3.9 ± 0.5 versus -6.4 ± 1.3; p = 0.004) and osteocalcin (mean ΔCt 2.7 ± 0.2 versus 1.8 ± 0.1; p = 0.004) (Fig. 2A). In the SaOS-2 cell line, overexpression of BMP-2 was correlated with an increase in the expression of Osterix (mean ΔCt -2.1 ± 0.6 versus -3.9 ± 0.6; p = 0.002) and a modest reduction in the expression of osteocalcin (mean ΔCt 2.8 ± 0.2 versus 3.1 ± 0.3; p = 0.02) (Fig. 2B). In the 143b cell line, BMP-2 upregulation was not correlated with a response or change in gene expression for any of the assessed markers (Fig. 2C). Lastly, in DLM8-M1 cells, the expression of BMP-2 was correlated with a reduction in the expression of connective tissue growth factor (mean ΔCt 0.2 ± 0.1 versus 0.6 ± 0.1; p = 0.002) and an increase in the expression of Runx-2 (mean ΔCt -0.8 ± 0.1 versus -1.1 ± 0.1; p = 0.002) (Fig. 2D). These results indicate that SaOS-2 is a more differentiated cell line than 143b with regards to the markers tested. Furthermore, the results indicate a lack of osteoblastic differentiation in 143b cells with BMP-2 overexpression and a partial differentiation response in SaOS-2 and DLM8-M1. No cell line demonstrated a full differentiation response.

Fig. 2.

Fig. 2

Open in a new tab

A-D BMP-2 induction of osteoblastic differentiation mRNA markers differs between the osteosarcoma cell lines. (A) SaOS-2 had lower expression of the earlier osteoblast lineage marker, connective tissue growth factor, with a relative higher expression of late osteoblast lineage markers, Runx-2, Osterix, and alkaline phosphatase than did 143b, suggesting increased osteoblastic differentiation. (B) In SaOS-2, the expression of BMP-2 resulted in an increased expression of Osterix and decreased expression of osteocalcin. (C) In 143b, no gene expression changes were observed with BMP-2 upregulation. (D) In DLM8-M1, the upregulation of BMP-2 resulted in a reduction of the expression of connective tissue growth factor and increase in the expression of Runx-2. The data represent five or six biologic replicates and technical robotic duplicates. Statistics were performed using the Mann-Whitney U-test. The corresponding letters represent statistical significance as follows: ap < 0.05; bp < 0.01.

BMP-2 Signaling Stimulates Local Tumor Growth in 143b- and DLM8-M1-derived Tumors, and Modestly Reduces Tumor Growth in SaOS-2-derived Tumors

Osteosarcoma developed in 21 of 21 mice injected with 143b cells and grew over the course of 4 weeks (Fig. 3A-B). The frequency of lung metastasis was not different between the treatment and control groups for either the rhBMP-2/ absorbable collagen sponge model or the BMP-2 overexpression model (metastases developed in 12 of 21 mice). A diagnosis of osteosarcoma was confirmed histologically (by identifying malignant osteoid production) and was classified as high-grade, poorly differentiated osteosarcoma (Fig. 3C). A histologic examination revealed physeal invasion (not shown) and microscopic pulmonary metastases (Fig. 3D).

Fig. 3.

Fig. 3

Open in a new tab

A-D The 143b xenograft model produced locally destructive tumors with a propensity for lung metastasis. (A) Bioluminescence was successfully used as a modality to measure the tumor burden during the 4-week experiment. (B) In the BMP-2 upregulation model (Control: n = 7; BMP-2: n = 6), bioluminescence scores climbed during the 4-week experiment, with no differences between the two groups. A diagnosis of osteosarcoma was confirmed histologically in both (C) primary (black arrow represents malignant osteoid deposition) and (D) metastatic tumor deposits in the lung (white arrow). For bioluminescence scores, a two-way analysis of variance was performed with a post-hoc Sidak’s multiple comparisons test. A color image accompanies the online version of this article.

At the experimental endpoint, the tumor volume was increased in the rhBMP-2 with absorbable collagen sponge cohort, versus absorbable collagen sponge control (495 ± 91.9 mm3 versus 1335 ± 102.7 mm3, mean difference 840.3 mm3 [95% CI 671.7 to 1009]; p < 0.0001) (Fig. 4A). The microCT analysis demonstrated the osteolytic nature of the 143b cell line. Three dimensional reconstructions were generated in triplicate, which revealed a larger degree of osseous destruction in the rhBMP-2 with absorbable collagen sponge cohort than in the control absorbable collagen sponge cohort (images not shown). Lung metastases developed in one of four mice in the control absorbable collagen sponge cohort, and two or four mice in the rhBMP-2 with absorbable collagen sponge cohort.

Fig. 4.

Fig. 4

Open in a new tab

A-D BMP-2 stimulates local in vivo tumor growth and osteolysis in 143b osteosarcoma cells. The tumor volume was larger in BMP-2 stimulated tumors, in both (A) the rhBMP-2/ absorbable collagen sponge experiment and (B) the BMP-2 upregulation model. (C) Visual comparison of a 143b hindlimb tumor with BMP-2 upregulation versus control. (D) Increased osteolysis was seen, with an increased tumor burden in BMP-2-stimulated tumors. For tumor volume, an unpaired t-test was performed. A color image accompanies the online version of this article.

In BMP-2-overexpressed 143b tumors, bioluminescence scores consistently increased during the 4 weeks, with no difference in average bioluminescence values (p = 0.08) (Fig. 3B). At the experimental endpoint, the tumor volume was increased in the BMP-2 overexpression cohort (Fig. 4B-C) versus the control cell line (307.2 ± 106.8 mm3 versus 1316 ± 387.4 mm3, mean difference 1009 mm3 [95% CI 674.5 to 1343]; p < 0.001). The microCT analysis further demonstrated the osteolytic nature of the 143b cell line. Three dimensional reconstructions were generated in triplicate, which revealed a qualitatively larger degree of osseous destruction in the BMP-2 overexpression cohort than in control tumors (Fig. 4D). Lung metastases developed in five of seven mice in the control cohort versus four of six in the BMP-2 overexpression cohort. Overall, the results indicated that BMP-2, whether exogenously applied or ectopically expressed by osteosarcoma cell lines, resulted in an overall increase in the local tumor burden in the 143b cell line, but there was no difference in the frequency of lung metastases.

Osteosarcoma developed in 17 of 24 mice injected with SaOS-2 cells. None of those 17 mice developed lung metastasis. Tumor growth was not consistently palpable; therefore, we could not calculate the tumor volume. A diagnosis of high-grade, osteoblastic, intramedullary osteosarcoma was confirmed on histologic examination, and abundant, malignant osteoid deposition was identified (Fig. 5A). Tumors were monitored via bioluminescence over the course of 6 weeks (Fig. 5B). In the post-hoc analysis, the overexpression of BMP-2 reduced the tumor burden in SaOS-2 tumors at 6 weeks (9.5 x 108 ± 8.3 x 108 photons/s versus 9.3 x 107 ± 1.5 x 108 photons/s, mean difference 8.6 x 108 photons/s [95% CI 5.1 x 108 to 1.2 x 109]; p < 0.001) (Fig. 5B). Furthermore, the overexpression of BMP-2 resulted in decreased tumor-associated matrix deposition, as assessed by the qualitative and quantitative microCT analyses (Fig. 5C-D). The SaOS-2 control group of tumors had higher values for trabecular thickness (% change from contralateral extremity, mm) at the experimental endpoint (132.4 ± 8.4 versus 102.8 ± 3.6; p = 0.02) (Fig. 5C). There were no differences between groups with regard to trabecular number, trabecular spacing, trabecular connectivity density, or bone mineral density. Overall, these results suggest that increased BMP-2 partially impeded local tumor growth in the SaOS-2 cell line and had no effect on the frequency of lung metastases.

Fig. 5.

Fig. 5

Open in a new tab

A-D The overexpression of BMP-2 delayed the local in vivo tumor burden in SaOS-2 cells. (A) Tumors derived from SaOS-2 cells were histologically confirmed as osteosarcoma, demonstrating malignant osteoid deposition (arrow). Tumors were allowed to grow during a 6-week period and monitored using bioluminescence (Control: n = 8; BMP-2: n = 7). A reduction in tumor burden was seen in (B) BMP-2-stimulated tumors (p < 0.001). Tumor burden was further assessed using microCT, which was performed on seven tumors from the control cohort and five tumors from the BMP-2-overexpressing cohort. (C) Control tumors has an increased trabecular thickness compared with the BMP-2 cohort (p = 0.02), suggesting an increased tumor burden in the control cohort. (D) For bioluminescence data, a two-way ANOVA was performed with a post-hoc Sidak’s multiple comparisons test. For microCT data, statistics were performed using an unpaired t-test. A color image accompanies the online version of this article.

Osteosarcoma developed in eight of the 10 C3H mice injected with DLM8-M1 cells. Tumors developed in three of the five mice injected with DLM8-M1 control cells, while tumors developed in all five of mice with DLM8-M1-BMP-2 cells. There was a histologic diagnosis of a poorly differentiated osteosarcoma in local and metastatic tumors (Fig. 6A-B). One of three mice with DLM8-M1 control tumors developed lung metastases, and 2 of 5 mice with DLM8-M1-BMP-2-derived tumors developed lung metastases. When the bioluminescence signal was lost in vivo, tumor volume was used as a primary method for monitoring tumor growth. Tumor volume was larger in the BMP-2 cohort than in the control cohort [week 4: 0 mm3 versus 326.1 ± 72.8 mm3, mean difference 326.1 mm3 [95% CI 121.2 to 531]; p = 0.009 (Fig. 6C)]. By week 5, three mice in the control cohort developed palpable tumors, with a mean tumor volume of 120 ± 28 mm3. Before week 5, the BMP-2 cohort reached the humane endpoint. MicroCT facilitated a qualitative assessment of tumors, which demonstrated an increased tumor burden in the BMP-2 cohort (Fig. 6D). Overall, these results indicate that although the overexpression of BMP-2 stimulated an increase in orthotopic DLM8-M1 tumor growth in the tibia, it had no effect on the frequency of lung metastases.

Fig. 6.

Fig. 6

Open in a new tab

The overexpression of BMP-2 stimulated local in vivo tumor growth in the metastatic murine osteosarcoma cell line, DLM8-M1. (A) Local and (B) pulmonary metastatic tumors derived from DLM8-M1 cells were histologically confirmed as osteosarcoma (arrows represent malignant osteoid deposition). Tumors were allowed to grow until the experimental endpoint (5 weeks for the BMP-2 cohort; 6 weeks for the control group) (n = 3 for control; n = 5 for the BMP-2 cohort), with (C) an increased tumor volume in BMP-2-upregulating tumors (p = 0.009 at 5 weeks). (D) The tumor burden was qualitatively assessed using microCT, demonstrating increased burden in the BMP-2 cohort and a more-osteoblastic pattern than in 143b tumors. For tumor volume, an unpaired t-test was performed. The bioluminescence signal was lost after in vivo growth. A color image accompanies the online version of this article.

Discussion

Tissue healing in osteosarcoma limb salvage surgery is impaired due to use of cytotoxic chemotherapy and the presence of large bone and soft tissue defects. Osteosarcoma patients are therefore at risk of complications after allograft augmented limb reconstruction, such as nonunion, implant loosening, and infection. rhBMP-2 augmentation may improve bone healing [12], particularly in allograft-based limb salvage surgery; however, concerns regarding prooncogenic signaling has raised concerns about its clinical use in osteosarcoma surgery. Through a series of in vitro and in vivo studies, we assessed whether BMP-2 differentiates osteosarcoma cells, and if ectopic BMP-2 signaling affects local or systemic osteosarcoma tumor burden. Presence of ectopic BMP-2 signaling stimulated local tumor growth in two of three models (143b and DLM8-M1), and partially impeded growth in one model (SaOS-2). BMP-2 mediated osteoblastic differentiation was more impaired in the models where BMP-2 stimulated tumor growth (143b and DLM8-M1). These results support the current recommendation to avoid rhBMP-2 use in osteosarcoma limb-salvage surgery, as there remains potential that this osteoinductive growth factor may stimulate tumor growth of any residual microscopic disease.

Limitations

There are limitations to this study. Immortalized cell lines that have undergone in vitro passaging do not represent primary human clinical tissue samples because of genetic drift and homogeneity [10]. To mitigate this limitation, we included three osteosarcoma cell lines representing different degrees of osteoblastic differentiation. The orthotopic murine models used strives to recapitulate the human disease process but may not be relevant to human tumors. Constant BMP-2 delivery to the tumor bed via constitutive expression from osteosarcoma cell is an artificial representation of the clinical scenario of rhBMP-2 delivery to a sarcoma resection site. Our results using the rhBMP-2-soaked (weight-adjusted dose) absorbable collagen sponge in the 143b cell line, however, demonstrated consistent findings to the BMP-2 overexpression model, lending support to its use as a translational model. In terms of the method of assessing the tumor burden, we used tumor volume, bioluminescence, and microCT. MicroCT has not been previously rigorously assessed to quantify osteosarcoma tumor burden but provided valuable adjunctive qualitative and quantitative data to our assessment. In the 143b and DLM8-M1 models, the degree of osteolysis made our methods of segmentation and analysis highly variable and inaccurate, which did not allow for an accurate quantification of osteolysis; however, the degree of visualized bone loss was qualitatively apparent on 3-D reconstruction.

Osteosarcoma Differentiation and BMP-2 Signaling

Our results suggest a differential impact of BMP-2 signaling on the biology of osteosarcoma, whereby BMP-2 has a proliferative in vivo effect on 143b and DLM8-M1 cell lines while reducing proliferation in SaOS-2. Our results with the 143b tumors are consistent with those of others who demonstrated that osteogenic BMPs were unable to induce differentiation of certain osteosarcoma cells but could induce the growth of some cell lines [14, 17].

Our qRT-PCR results indicated that the overexpression of BMP-2 in 143b cells did not lead to a difference in the expression of the osteoblastic lineage markers connective tissue growth factor, Runx-2, Osterix, alkaline phosphatase, or osteocalcin. One study demonstrated that epidermal growth factor and fibroblast growth factor-2 signaling inhibited BMP-2-mediated osteoblastic differentiation via activation of the RTK-Ras-ERK pathway [20]. Oncogenic K-ras signaling in the 143b cell line may contribute to the lack of osteoblastic differentiation in response to BMP-2. Although Ras mutations are infrequent in osteosarcoma, activation of the downstream pathway (Raf/MEK/ERK) is present in osteosarcoma and is a candidate pathway for targeted inhibition of those cells [6].

Another study demonstrated that Coleusin factor induces a dose-dependent increase in the expression of BMP-2, which is correlated with an overexpression of differentiation markers (alkaline phosphatase, Runx-2, osteopontin, and osteocalcin) and a reduction in cell proliferation [9]. These results are consistent with our findings with SaOS-2 cells, suggesting that BMP-2 stimulation does not promote tumor growth; instead, it induces a more-differentiated phenotype. Our qRT-PCR results indicated that BMP-2 signaling was correlated with upregulation of Osterix. Our results are also similar to a study [5] that demonstrated the expression of Osterix could diminish the proliferative capacity of osteosarcoma in vitro and in vivo and reduce lung metastases in an orthotopic mouse model. Interestingly, BMP-2 induced a minor differentiation response in DLM8-M1 (reduction in the expression of connective tissue growth factor and upregulation of Runx-2) but was still able to stimulate tumor proliferation in vivo. The relationship between Runx-2 and osteosarcoma is complex. For example, Runx-2 may induce a differentiation response and inhibition of growth in osteosarcoma, [26], but overexpression of Runx-2 in osteosarcoma has also been correlated with chemotherapy resistance [24]. It is plausible that in DLM8-M1 cells, BMP-2 does not stimulate differentiation past the Runx-2 “checkpoint” and may promote tumor growth via stimulation of mitogenic pathways. Notably, osteocalcin appears to be downregulated in SaOS-2 cells compared with 143b cells and after stimulation by BMP-2. Most models favor osteocalcin as a marker of terminal differentiation; however, its relationship with OS differentiation is less clear. The gene expression of osteocalcin did not appear to be correlated with the osteoblastic phenotype, as suggested by our microCT analysis.

Translational Impact

Because clinical osteosarcoma samples represent a phenotypically diverse spectrum of cancers, this variable response to BMP-2 signaling could be clinically recapitulated in different patients with osteosarcoma. As discussed by Weiss [28] in a narrative review, although applying a growth factor such as rhBMP-2 on an absorbable collagen sponge to a tumor bed may improve local outcomes and induce differentiation, it may also introduce the risk of potentiating tumor growth and metastasis. Furthermore, there is no clinical evidence to suggest that rhBMP-2 would improve allograft union in this setting, although it might be beneficial [28]. Without rigorous data showing that rhBMP-2 on an absorbable collagen sponge is absolutely not prooncogenic in osteosarcoma, its clinical use should remain limited in an effort to do no harm. With our current knowledge of the potentially protumor biologic response to BMP-2 signaling and the data obtained from this study, we suggest that rhBMP-2 on an absorbable collagen sponge should continue to be avoided in limb-salvage surgery for osteosarcoma.

Future Directions

Future directions for research include the addition of more human osteosarcoma cell lines along the spectrum of osteoblastic differentiation. Because BMP-2 led to a decreased tumor burden in a more differentiated osteosarcoma model, further investigation into how BMP-2 may be coupled with other established osteosarcoma differentiation therapies is warranted. Furthermore, assessing the orthobiologic function of these combined therapies may augment bone healing. Because the differentiation profile is correlated with the in vivo phenotype, future research might assess tumor differentiation using a spectrum of osteoblastic markers that correlate with the clinical phenotype and prognosis.

Conclusions and Clinical Relevance

In orthotopic murine models of osteosarcoma, BMP-2 stimulated local tumor growth of 143b and DLM8-M1-derived tumors but partially limited the tumor burden in SaOS-2-derived tumors. BMP-2 had no effect on the frequency of pulmonary metastases in any cell line model. As the potential for BMP-2 mediated osteoblastic differentiation of osteosarcoma cells is not reliably assessed in clinical samples, BMP-2 may have unpredictable oncologic consequences on microscopic residual tumor cells if used in limb-salvage surgery to treat osteosarcoma. Our study provides further support for the recommendation that BMP-2 should continue to be avoided at the time of the index limb-salvage surgery in osteosarcoma. Further investigations into the role of BMP-2-mediated osteosarcoma differentiation and pathway augmentation are warranted.

Acknowledgments

We thank the staff of the University of Calgary Bone Imaging Laboratory for their assistance with using the microCT equipment, as well as Franz Zemp PhD, for contributing transfection reagents. We also thank Teresa Sheidl-Yee MSc, for her assistance in planning the molecular cloning experiments.

Footnotes

The institution of one or more of the authors (MJM, JKK) has received, during the study period, funding from the Canadian Orthopaedic Foundation (2017 Robert B. Salter Award).

Each author certifies that neither he or she, nor any member of his or her immediate family, has funding or commercial associations (consultancies, stock ownership, equity interest, patent/licensing arrangements, etc) that might pose a conflict of interest in connection with the submitted article.

All ICMJE Conflict of Interest Forms for authors and Clinical Orthopaedics and Related Research® editors and board members are on file with the publication and can be viewed on request.

Clinical Orthopaedics and Related Research® neither advocates nor endorses the use of any treatment, drug, or device. Readers are encouraged to always seek additional information, including FDA approval status, of any drug or device before clinical use.

Each author certifies that his or her institution approved the animal protocol for this investigation and that all investigations were conducted in conformity with ethical principles of research.

References

  • 1.Aanstoos ME, Regan DP, Rose RJ, Chubb LS, Ehrhart NP. Do mesenchymal stromal cells influence microscopic residual or metastatic osteosarcoma in a murine model? Clin Orthop Relat Res . 2016;474:707-715. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Asai T, Ueda T, Itoh K, Yoshioka K, Aoki Y, Mori S, Yoshikawa H. Establishment and characterization of a murine osteosarcoma cell line (LM8) with high metastatic potential to the lung. Int J Cancer. 1998;76:418-422. [DOI] [PubMed] [Google Scholar]
  • 3.Bishop GB, Einhorn TA. Current and future clinical applications of bone morphogenetic proteins in orthopaedic trauma surgery. Int Orthop. 2007;31:721-727. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Bondareva A, Downey CM, Ayres F, Liu W, Boyd SK, Hallgrimsson B, Jirik FR. The lysyl oxidase inhibitor, beta-aminopropionitrile, diminishes the metastatic colonization potential of circulating breast cancer cells. PLoS One. 2009;4:e5620. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Cao Y, Zhou Z, de Crombrugghe B, Nakashima K, Guan H, Duan X, Jia S, Kleinerman ES. Osterix, a transcription factor for osteoblast differentiation, mediates antitumor activity in murine osteosarcoma. Cancer Res. 2005;65:1124–1128. [DOI] [PubMed] [Google Scholar]
  • 6.Chandhanayingyong C, Kim Y, Staples JR, Hahn C, Lee FY. MAPK/ERK Signaling in Osteosarcomas, Ewing Sarcomas and Chondrosarcomas: Therapeutic Implications and Future Directions. Sarcoma . 2012;2012:404810. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Cheng H, Jiang W, Phillips FM, Haydon RC, Peng Y, Zhou L, Luu HH, An N, Breyer B, Vanichakarn P, Szatkowski JP, Park JY, He T. Osteogenic activity of the fourteen types of human bone morphogenetic proteins (BMPs). J Bone Joint Surg Am. 2003;85:1544–1552. [DOI] [PubMed] [Google Scholar]
  • 8.Geller DS, Singh MY, Zhang W, Gill J, Roth ME, Kim MY, Xie X, Singh CK, Dorfman HD, Villanueva-Siles E, Park A, Piperdi S, Gorlick R. Development of a Model System to Evaluate Local Recurrence in Osteosarcoma and Assessment of the Effects of Bone Morphogenetic Protein-2. Clin Cancer Res . 2015;21:3003–3012. [DOI] [PubMed] [Google Scholar]
  • 9.Geng S, Sun B, Lu R, Wang J. Coleusin factor, a novel anticancer diterpenoid, inhibits osteosarcoma growth by inducing bone morphogenetic protein-2-dependent differentiation. Mol Cancer Ther. 2014;13:1431–1441. [DOI] [PubMed] [Google Scholar]
  • 10.Gengenbacher N, Singhal M, Augustin HG. Preclinical mouse solid tumour models: status quo, challenges and perspectives. Nat Rev Cancer. 2017;17:751–765. [DOI] [PubMed] [Google Scholar]
  • 11.Gill J, Connolly P, Roth M, Chung SH, Zhang W, Piperdi S, Hoang B, Yang R, Guzik H, Morris J, Gorlick R, Geller DS. The effect of bone morphogenetic protein-2 on osteosarcoma metastasis. PLoS One . 2017;12:e0173322. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Govender S, Csimma C, Genant HK, Valentin-Opran A, Amit Y, Arbel R, Aro H, Atar D, Bishay M, Börner MG, Chiron P, Choong P, Cinats J, Courtenay B, Feibel R, Geulette B, Gravel C, Haas N, Raschke M, Hammacher E, van der Velde D, Hardy P, Holt M, Josten C, Ketterl RL, Lindique B, Lob G, Mathevon H, McCoy G, Marsh D, Miller R, Munting E, Oevre S, Nordsletten L, Patel A, Pohl A, Rennie W, Reynders P, Rommens PM, Rondia J, Rossouw WC, Daneel PJ, Ruff S, Rüter A, Santavirta S, Schildhauer TA, Gekle C, Schnettler R, Segal D, Seiler H, Snowdowne RB, Stapert J, Taglang G, Verdonk R, Vogels L, Weckback A, Wentzensen A, Wisniewski T, BMP-2 Evaluation in Surgery for Tibial Trauma (BESTT) Study Group. Recombinant human bone morphogenetic protein-2 for treatment of open tibial fractures: a prospective, controlled, randomized study of four hundred and fifty patients. J Bone Joint Surg Am. 2002;84:2123-2134. [DOI] [PubMed] [Google Scholar]
  • 13.Grimer RJ. Surgical options for children with osteosarcoma. Lancet Oncol. 2005;6:85–92. [DOI] [PubMed] [Google Scholar]
  • 14.Haydon RC, Luu HH, He T-C. Osteosarcoma and osteoblastic differentiation: a new perspective on oncogenesis. Clin Orthop Relat Res . 2007;454:237–246. [DOI] [PubMed] [Google Scholar]
  • 15.James AW, LaChaud G, Shen J, Asatrian G, Nguyen V, Zhang X, Ting K, Soo C. A Review of the Clinical Side Effects of Bone Morphogenetic Protein-2. Tissue Eng Part B Rev. 2016;22:284–297 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Jeys LM, Kulkarni A, Grimer RJ, Carter SR, Tillman RM, Abudu A. Endoprosthetic reconstruction for the treatment of musculoskeletal tumors of the appendicular skeleton and pelvis. J Bone Joint Surg Am . 2008;90:1265–1271. [DOI] [PubMed] [Google Scholar]
  • 17.Luo X, Chen J, Song W-X, Tang N, Luo J, Deng Z-L, Sharff KA, He G, Bi Y, He B-C, Bennett E, Huang J, Kang Q, Jiang W, Su Y, Zhu G-H, Yin H, He Y, Wang Y, Souris JS, Chen L, Zuo G-W, Montag AG, Reid RR, Haydon RC, Luu HH, He T-C. Osteogenic BMPs promote tumor growth of human osteosarcomas that harbor differentiation defects. Lab Invest . 2008;88:1264–1277. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Luu HH, Kang Q, Park JK, Si W, Luo Q, Jiang W, Yin H, Montag AG, Simon MA, Peabody TD, Haydon RC, Rinker-Schaeffer CW, He T-C. An orthotopic model of human osteosarcoma growth and spontaneous pulmonary metastasis. Clin Exp Metastasis. 2005;22:319–329. [DOI] [PubMed] [Google Scholar]
  • 19.Morris CD, Wustrack RL, Levin AS. Limb-Salvage Options in Growing Children with Malignant Bone Tumors of the Lower Extremity: A Critical Analysis Review. JBJS Rev. 2017;5:e7. [DOI] [PubMed] [Google Scholar]
  • 20.Nakayama K, Tamura Y, Suzawa M, Harada S, Fukumoto S, Kato M, Miyazono K, Rodan GA, Takeuchi Y, Fujita T. Receptor tyrosine kinases inhibit bone morphogenetic protein-Smad responsive promoter activity and differentiation of murine MC3T3-E1 osteoblast-like cells. J Bone Miner Res. 2003;18:827–835. [DOI] [PubMed] [Google Scholar]
  • 21.Pensak M, Hong S, Dukas A, Tinsley B, Drissi H, Tang A, Cote M, Sugiyama O, Lichtler A, Rowe D, Lieberman JR. The role of transduced bone marrow cells overexpressing BMP-2 in healing critical-sized defects in a mouse femur. Gene Ther . 2015;22:467–475 [DOI] [PubMed] [Google Scholar]
  • 22.Rhim JS, Cho HY, Huebner RJ. Non-producer human cells induced by murine sarcoma virus. Int J Cancer . 1975;15:23–29. [DOI] [PubMed] [Google Scholar]
  • 23.Rosen V. BMP2 signaling in bone development and repair. Cytokine Growth Factor Rev. 2009;20:475–480 [DOI] [PubMed] [Google Scholar]
  • 24.Sadikovic B, Thorner P, Chilton-Macneill S, Martin JW, Cervigne NK, Squire J, Zielenska M. Expression analysis of genes associated with human osteosarcoma tumors shows correlation of RUNX2 overexpression with poor response to chemotherapy. BMC Cancer. 2010;10:202. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Singla AK, Downey CM, Bebb GD, Jirik FR. Characterization of a murine model of metastatic human non-small cell lung cancer and effect of CXCR4 inhibition on the growth of metastases. Oncoscience . 2015;2:263-271. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Smida J, Baumhoer D, Rosemann M, Walch A, Bielack S, Poremba C, Remberger K, Korsching E, Scheurlen W, Dierkes C, Burdach S, Jundt G, Atkinson MJ, Nathrath M. Genomic alterations and allelic imbalances are strong prognostic predictors in osteosarcoma. Clin Cancer Res. 2010;16:4256–4267. [DOI] [PubMed] [Google Scholar]
  • 27.Wang L, Park P, Zhang H, La Marca F, Claeson A, Valdivia J, Lin C. BMP-2 inhibits the tumorigenicity of cancer stem cells in human osteosarcoma OS99-1 cell line. Cancer Biol Ther. 2011;11:457-463. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Weiss KR. “To B(MP-2) or not to B(MP-2)” or “Much Ado About Nothing”: Are orthobiologics in tumor surgery worth the risks? Clin Cancer Res. 2015;21:2889–2891. [DOI] [PubMed] [Google Scholar]
  • 29.Yu YY, Lieu S, Lu C, Colnot C. Bone morphogenetic protein 2 stimulates endochondral ossification by regulating periosteal cell fate during bone repair. Bone. 2010;47:65–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Yuan J, Ossendorf C, Szatkowski JP, Bronk JT, Maran A, Yaszemski M, Bolander ME, Sarkar G, Fuchs B. Osteoblastic and osteolytic human osteosarcomas can be studied with a new xenograft mouse model producing spontaneous metastases. Cancer Invest . 2009;27:435–442. [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from Clinical Orthopaedics and Related Research are provided here courtesy of The Association of Bone and Joint Surgeons

RESOURCES